摘要
CT和MRI图像断层之间的距离远大于断层内部像素间的距离,三维剂量场的计算等工作通常需要等间隔分布的三维图像数据。目前常用的基于灰度插值方法会引起图像边界模糊,而基于形状的插值方法不能得到整个图像的数据。为解决这一问题,文中提出了一种基于断层图像分割的三维匹配插值算法。通过对断层图像进行分割,获得断层图像的空气、软组织和骨骼等区域信息。对相同密度区域采用匹配插值,不同密度区域采用缩放区域大小作为插值数据,使新的图像不仅在灰度上,而且在组织形状上介于原来的断层图像之间,满足了医学图像插值要求。和线性插值方法相比,新算法提高了插值图像的质量,插值结果可有效地应用于构建三维体模型。
Dose calculation in radiotherapy needs equal spacing3D images data sets,which can be reconstructed from a sequence of cross-sectional slices(such as CT,MRI etc).Typically,the spacing between slices is greater than the spacing between points on a slice.To get equal sample spacing in all directions,one usually uses grey-baed interpola-tion method.Unfortunately,grey-baed interpolation makes the images blurred on edge.Another usually uses shape-based interpolation that doesn't get the data of whole image.In order to solve the problem,the paper introduces a new interpo-lation algorithm based on images segmentation.Firstly,the algorithm obtains the area of air,soft tissue and skeleton through segmenting images.Then,the algorithm uses matching interpolation in the same density areas and scales the size of area as the interpolation data in the different density area.So that the new image basically satisfies the requirements of medical image interpolation.Compared with linear interpolation,the new algorithm improves the quality of image.The interpolation can be effectively used to construct3D volume models.
出处
《计算机工程与应用》
CSCD
北大核心
2004年第24期80-82,共3页
Computer Engineering and Applications
关键词
匹配插值
三维图像
门限分割
体模型
matching interpolation,3-D image,threshold segmentation,volume model
